Enregistré dans:
Détails bibliographiques
Auteurs principaux: Wang, Yinuo, Wang, Likun, Jiang, Yuxuan, Zou, Wenjun, Liu, Tong, Song, Xujie, Wang, Wenxuan, Xiao, Liming, Wu, Jiang, Duan, Jingliang, Li, Shengbo Eben
Format: Preprint
Publié: 2024
Sujets:
Accès en ligne:https://arxiv.org/abs/2405.15177
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866910757701025792
author Wang, Yinuo
Wang, Likun
Jiang, Yuxuan
Zou, Wenjun
Liu, Tong
Song, Xujie
Wang, Wenxuan
Xiao, Liming
Wu, Jiang
Duan, Jingliang
Li, Shengbo Eben
author_facet Wang, Yinuo
Wang, Likun
Jiang, Yuxuan
Zou, Wenjun
Liu, Tong
Song, Xujie
Wang, Wenxuan
Xiao, Liming
Wu, Jiang
Duan, Jingliang
Li, Shengbo Eben
contents Reinforcement learning (RL) has proven highly effective in addressing complex decision-making and control tasks. However, in most traditional RL algorithms, the policy is typically parameterized as a diagonal Gaussian distribution with learned mean and variance, which constrains their capability to acquire complex policies. In response to this problem, we propose an online RL algorithm termed diffusion actor-critic with entropy regulator (DACER). This algorithm conceptualizes the reverse process of the diffusion model as a novel policy function and leverages the capability of the diffusion model to fit multimodal distributions, thereby enhancing the representational capacity of the policy. Since the distribution of the diffusion policy lacks an analytical expression, its entropy cannot be determined analytically. To mitigate this, we propose a method to estimate the entropy of the diffusion policy utilizing Gaussian mixture model. Building on the estimated entropy, we can learn a parameter $α$ that modulates the degree of exploration and exploitation. Parameter $α$ will be employed to adaptively regulate the variance of the added noise, which is applied to the action output by the diffusion model. Experimental trials on MuJoCo benchmarks and a multimodal task demonstrate that the DACER algorithm achieves state-of-the-art (SOTA) performance in most MuJoCo control tasks while exhibiting a stronger representational capacity of the diffusion policy.
format Preprint
id arxiv_https___arxiv_org_abs_2405_15177
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Diffusion Actor-Critic with Entropy Regulator
Wang, Yinuo
Wang, Likun
Jiang, Yuxuan
Zou, Wenjun
Liu, Tong
Song, Xujie
Wang, Wenxuan
Xiao, Liming
Wu, Jiang
Duan, Jingliang
Li, Shengbo Eben
Machine Learning
Artificial Intelligence
Reinforcement learning (RL) has proven highly effective in addressing complex decision-making and control tasks. However, in most traditional RL algorithms, the policy is typically parameterized as a diagonal Gaussian distribution with learned mean and variance, which constrains their capability to acquire complex policies. In response to this problem, we propose an online RL algorithm termed diffusion actor-critic with entropy regulator (DACER). This algorithm conceptualizes the reverse process of the diffusion model as a novel policy function and leverages the capability of the diffusion model to fit multimodal distributions, thereby enhancing the representational capacity of the policy. Since the distribution of the diffusion policy lacks an analytical expression, its entropy cannot be determined analytically. To mitigate this, we propose a method to estimate the entropy of the diffusion policy utilizing Gaussian mixture model. Building on the estimated entropy, we can learn a parameter $α$ that modulates the degree of exploration and exploitation. Parameter $α$ will be employed to adaptively regulate the variance of the added noise, which is applied to the action output by the diffusion model. Experimental trials on MuJoCo benchmarks and a multimodal task demonstrate that the DACER algorithm achieves state-of-the-art (SOTA) performance in most MuJoCo control tasks while exhibiting a stronger representational capacity of the diffusion policy.
title Diffusion Actor-Critic with Entropy Regulator
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2405.15177